[cwatson1982]

I took my g0704 apart to try to figure out some bad stick slip problems I was having on the Y axis (The ball screw has no measurable backlash with a .0005 DTI but total system error was .006! on the Y depending on the table position which made backlash comp kind of useless as it varied significantly across the travel). It turns out the the ways on the base are not coplanar at all. With the cross slide scraped flat and level I literally had .01 of diagonal rock on the Y (measured gap).

My options are:

scrape the cross slide to fit the base and live with the resulting table twist.
buy a straight edge casting and spend days trying to fix the base
replace the base and hope I get a better one.
pack it full of tannerite and see how tough cast iron is!

What would you do?

I kind of post this in hope that people who are completely new to this all like I was check these things before their warranty is up. If I had known how to check mine and done so this whole thing would have been returned ASAP but unfortunately I am about 3 months past the end of my warranty. Before I took it apart you could shine a flashlight through the right side of the base and y axis mating surfaces with the gib in place and see light on the other side. (from right to left)

[Winnfield]

That is awful. Honestly, I cannot see an easy way to fix that if you aren't good at scraping. Scraping a surface flat and fixing geometry problems on a dovetail are very different levels of skill. I'd take it to a machine shop and have them mill the dovetails off and pick up some linear rails on ebay. You can use a piece of Starrett ground flat stock from Enco or eBay as your saddle, and you'll have a cooler machine when you are done. it really stinks to have to deal with that, but there is a way to come out of it with something much better (linear rails).

[cwatson1982]

I may very well go that route to salvage things. I have a friend who can take the dovetails off on a VMC for me. Just have to sit down and try to figure out if I can adapt the ballscrew mounts I currently have. If I could figure out how to convert solely the Y axis for now I would be able to pull it off in a couple of days.

[Winnfield]

It's good to hear that you have access to someone who can do that for you. Good luck - I hope it works out!

[cwatson1982]

Anyone have any rough guesstimates for required load ratings on linear guides? It looks like I could fit 15mm miniature THK guides just by milling off the dovetails on the base and cross slide. I would still be able to keep the current ballscrew mounts due to their low height (about 16mm). They have a dynamic LR of about 1800 lbf and 2500 static which sounds a bit low. I could probably fit 20mm in there and maintain the current heights but I would have to remove more cast iron than I would like. As it is I'll likely stick the base and cross slide in my heat treat oven to stress relieve them post machining.

Things to consider:
I mainly use the mill to profile annealed steel: 1095, s35vn, 154cm, etc. with small cutters, 3/8 rougher being the absolute largest but usually 3/16ths to 1/4inch variable flute/helix carbide.

I'll likely be getting something significantly bigger in the next year or two (hoping to find a used robodrill or get a novakon/tormach/zx-45 retrofit if not) so this machine would end up relegated to cutting micarta, g10 and a little carbon fiber after that.

[wizard]

I'm a little late to this thread but you might want to consider having the ways milled or ground. It really sounds like you have a pretty significant error in the ways so the question really is: is there enough meat left to take this route?

That is one consideration, the other is the possibility that the casting has relieved some built in stress and that it has a way to go to further relieve that stress. Any machining could lead to further movement.

Finally you don't Mary the ways to the slide, rather you Mary the saddle to the ways. In otherwords you need to get the ways right before working on the saddle.

Anyone have any rough guesstimates for required load ratings on linear guides? It looks like I could fit 15mm miniature THK guides just by milling off the dovetails on the base and cross slide.

This would be somewhat dependent on your goals. However reference the manufactures documents and make sure the rail and car combo you select can handle loads in all directions.

I would still be able to keep the current ballscrew mounts due to their low height (about 16mm). They have a dynamic LR of about 1800 lbf and 2500 static which sounds a bit low. I could probably fit 20mm in there and maintain the current heights but I would have to remove more cast iron than I would like. As it is I'll likely stick the base and cross slide in my heat treat oven to stress relieve them post machining.

I could be wrong but I suspect you are focused on the wrong type of rail as 15mm is awfully thin. Again the rail you choose has to handle loads in all directions.

Things to consider:
I mainly use the mill to profile annealed steel: 1095, s35vn, 154cm, etc. with small cutters, 3/8 rougher being the absolute largest but usually 3/16ths to 1/4inch variable flute/helix carbide.

I'll likely be getting something significantly bigger in the next year or two (hoping to find a used robodrill or get a novakon/tormach/zx-45 retrofit if not) so this machine would end up relegated to cutting micarta, g10 and a little carbon fiber after that.

The mill still has to work properly and safely.

[cwatson1982]

Thanks, I would be using 2x rails and 4 blocks per axis but they may be too small.

The cost of having the ways reground would likely be higher than the cost of the mill unfortunately.

[Winnfield]

You will need HSR rails/trucks. 15mm would work for under the table, but for the saddle, you will need HSR20 or larger (hsr25). The base will need to be flat and perpendicular to the column, obviously, so your first goal ought to be ensuring that the plane you are milling for the rails is as close to perfectly perpendicular to the column as possible. If it were me doing it, I would scrape the base flat now before anything else. That way yo can use it to fixture it for milling. When you mill the dovetails away, you don't mill off then entire way, just the angled portion, then, if needed, you can step it up with Starrett ground 1018 flat stock. I use this stuff all the time for situations like this and it is great. If you mill away the entire dovetail, the casting WILL move. You can almost guarantee it You will lose some Z travel, but not much.

[Winnfield]

...Also, if you can find preloaded rails, you will be much happier with the result. The preload (radial clearance) is in the part number after the sealing/dustproofing designation and will be "C0" "C1" or "C2"

[cwatson1982]

Thanks for the info Winnfield. Seems like I have some planning to do! I was really trying to keep from having to re-make the y motor/bearing block and ballnut mount but if 15mm will be too small I'll just have to bite the bullet on those to one degree or another. I realize I didn't need to remove the entire dovetail but will have to mill a bit inward on the base to make room of the trucks to clear if using larger than 15mm. In the long run it may be easier to just buy a big hunk of steel for the base and start from scratch.

I can likely get the base flat, mill the dovetail off, stress relieve and then re-mill. I make knives for a living and have the equipment for the stress relieving part at least.

Being minus a little Z isn't too big of a deal for what I do.

I think I'll probably purchase another base and if it's better than the one I have now - use it on the mill while I get the worse of the two ready for rails. For profiling parts the table twist I have now isn't huge since they have to be re-finished by hand anyway. The stiction is what kills me with quadrant marks due to it being significantly variable across the travel.

[Winnfield]

A large milling vise is a good platform for a mill with linear rails. You can get an 8" width vise for a few hundred, and the bed is ground flat, ready for rails.